irsvm function

fit case weighted support vector machines with robust loss functions

Fit case weighted support vector machines with robust loss functions. This is the wrapper function of irsvm_fit, which does the computing.

## S3 method for class 'formula'
irsvm(formula, data, weights, contrasts=NULL, ...)
## S3 method for class 'matrix'
irsvm(x, y, weights, ...)
## Default S3 method:
irsvm(x,  ...)

Arguments

• formula: symbolic description of the model, see details.
• data: argument controlling formula processing via model.frame.
• weights: optional numeric vector of weights
• x: input matrix, of dimension nobs x nvars; each row is an observation vector
• y: response variable. Quantitative for type="eps-regression", "nu-regression" and -1/1 for type="C-classification", "nu-Classification".
• contrasts: the contrasts corresponding to levels from the respective models
• ...: Other arguments passing to irsvm_fit

Details

Fit a robust SVM where the loss function is a composite function cfunotype + penalty. The model is fit by the iteratively reweighted SVM, an application of the iteratively reweighted convex optimization (IRCO). Here convex is the loss function induced by type.

For linear kernel, the coefficients of the regression/decision hyperplane can be extracted using the coef method.

Returns

An object with S3 class "wsvm" for various types of models. - call: the call that produced this object

• weights_update: weights in the final iteration of the IRCO algorithm

• cfun, s: original input arguments

• delta: delta value used for cfun="gcave"

References

Zhu Wang (2024) Unified Robust Estimation, Australian & New Zealand Journal of Statistics. 66(1):77-102.

Author(s)

Zhu Wang zwang145@uthsc.edu

See Also

irsvm_fit, print, predict, coef.

Examples

#binomial
x=matrix(rnorm(100*20),100,20)
g2=sample(c(-1,1),100,replace=TRUE)
fit=irsvm(x,g2,s=1,cfun="ccave",type="C-classification")